SaaS Formulas Are Easy to Find—Knowing Which Numbers Lie to You Is Hard
Deep thoughts on AI and aspirations —— ByteDance Deep Thinking Circle
SaaS metric formulas are searchable online: how to calculate LTV, how to calculate CAC, how to calculate net revenue retention. But knowing how to calculate and knowing how to read are two different things. The same number tells completely different stories at different stages and under different definitions. A seemingly healthy metric may hide the exact opposite conclusion underneath. This article doesn’t list formulas—it focuses on the judgment points that are easiest to misread.
Three Numbers Most Easily Misread
The first is the LTV to CAC ratio. The rule of thumb is that greater than 3 is healthy, and most people stop there, remembering only “higher is better.” But there’s a second half to this ratio: too high actually indicates wasted opportunity. An absurdly high LTV:CAC often means severely insufficient acquisition investment—you could be spending more money to buy growth, but you’re saving it instead. Saving isn’t skill; not spending the money is the problem.
The second is payback period. Looking only at the ratio without considering payback period is a common starting point for cash flow disasters. No matter how high the LTV, it’s recovered over the entire customer lifetime. If recovering full acquisition costs takes two years, you’re booking a large loss upfront with every deal signed. The faster you grow, the bigger the cash gap. Subscription businesses typically consider recovery within one year as the health threshold. Beyond this line, expansion basically requires external capital. So whether an acquisition engine is healthy requires looking at at least three numbers together: the ratio, payback months, and how long cash on hand can last.
The third is net revenue retention. This metric answers the question: ignoring acquisition entirely, does revenue from existing customers rise or fall? Expansions from existing customers minus contractions and churn—what’s left. Above 100% means even if you freeze the sales team, revenue continues growing naturally. Hitting 120% is equivalent to getting 20% growth for free each year. But it has a blind spot: it only tracks changes in cohort customers between period start and end. New customer contributions are completely excluded. A company that relies on new customers to pump revenue while existing customers continuously contract can still make this metric look good.
More error-prone than formulas is actually definition. Retention measured by customer count versus revenue amount tells two different stories: the former reflects product stickiness, the latter reflects whether money stayed. Net revenue retention and gross revenue retention must also be viewed separately—one includes expansion, the other only measures what remains of original revenue. Once definitions get mixed up, the same business can yield two completely opposite conclusions of “very healthy” and “very dangerous.” Revenue itself is the same: you need to break down the structure. New, expansion, contraction, and churn—these four components add up to total growth. A company growing through expansion versus one growing through new acquisition have completely different playbooks for the next step.
Single Metrics Deceive—Cross-Checks Reveal Judgment
Metrics need to check each other.
The most classic pairing is the tradeoff between growth and profit. The industry’s Rule of 40: revenue growth rate plus profit margin—the sum cannot fall below 40%. The logic is straightforward: you need to lead in either growth or profit. Capital markets pay a premium for either growth or profit. Companies weak in both get compressed valuation multiples. This rule has boundaries—it targets companies that have reached a certain scale. Applying it to early-stage companies still finding direction is meaningless.
Another pairing is churn and growth cross-checks. Surface-level rapid growth with monthly churn approaching the 5% health warning line equals pouring water into a leaking bucket. Churn rate also cannibalizes lifetime value through the intermediary variable of lifecycle: 3% monthly churn means average customer lifetime of only about 33 months, and all lifetime value estimates shrink accordingly. When looking at growth, always ask simultaneously: what’s the churn?
When placing several metrics together, the questions each answers and common misreadings look like this:
| Metric | Question Answered | Where It’s Easily Misread |
|---|---|---|
| LTV:CAC | Is the acquisition engine profitable | Too high = insufficient investment, not the higher the better |
| Payback Period | Does cash flow turn | Good ratio but slow recovery still kills you |
| Net Revenue Retention | Is the existing base expanding or leaking | Excludes new customers; can’t detect facade propped up by new business |
| Growth + Profit | Does valuation logic hold | Doesn’t apply to early stage; don’t force it |
| Monthly Churn Rate | Is the bottom leaking | Fast growth masks leakage; must view separately |
Metrics Follow Stage, Not Dashboard
For the same set of metrics, which ones to watch at different stages yields different answers. During product-market fit, focus on retention, activation, and core feature usage. Acquisition cost is meaningless at this point because you haven’t started serious acquisition. During validation of repeatable growth, the protagonists become acquisition cost, payback period, and lead conversion. Entering scaled operations, it’s the combination of growth plus profit and cost structure’s turn. Cost structure has established experience ranges for comparison: marketing and sales costs shouldn’t exceed 40% of revenue, service and R&D each occupy one to two tenths, administration compressed under 10%. Exceeding ranges requires organizational inspection; falling far below ranges requires suspecting whether money was saved where it should have been spent.
Keep metric count low, but each should answer an action. Seeing a number deteriorate should immediately indicate what to check next: activation, service, or pricing. If nobody knows what to do after a number changes, it’s noise. Keeping it on reports only dilutes attention.
The worst practice is treating metrics as performance evaluation rather than diagnostic tools. Once renewal rate becomes the sole KPI for customer success teams, teams optimize around this number: suppressing prices, pleasing customers, postponing churn. The number looks good, but the business may be worse. Once metrics become performance targets, they lose their alarm function.
Use Metrics as Business Assumption Translators
I habitually use metrics as translators for business assumptions. First write down the most worrying question right now: acquisition too expensive, can’t retain, or growth is hollow. Then select two or three metrics that answer this question, provide an explanation, set an action, and schedule the next review. When numbers change, first ask “is my explanation still correct,” then decide what to change.
Formulas don’t change; the reading does. The reading differs at every stage. Many people can recite formulas; few know which number to trust at this moment. That latter ability is what’s called business capability.